Abstract of Supplement Alzheimer's disease (AD) is a devastating condition affecting millions and costing hundreds of billions of dollars in caregiving expenses, for which no effective therapy exists. The pathophysiology of AD is complex and not fully understood, but nutrition and exercise represent the most significant lifestyle determinants of AD risk. Mitochondria have been linked to neurodegeneration on many levels; however, their DNA has not been carefully studied in the context of Alzheimer's disease risk. Our lab characterized humanin, a mitochondrial-derived peptide encoded from the 16S rRNA region of the mtDNA, which has shown to be a potent neuroprotective factor. We recently showed that humanin administration prevents age-dependent cognitive decline in mice, and others have shown that it slows the progression of amyloid beta (A?)-driven neurodegeneration in AD mouse models. In recent unpublished data presented in this grant, separately, we recently identified a SNP in the humanin ORF that is associated with lower circulating humanin levels and with more rapid cognitive decline in individuals followed in the Health and Retirement Study (HRS). We now propose to expand this paradigm to the entire mitochondrial genome to discover novel peptides that modulate AD and purpose the hypotheses that: 1) mitochondrial genetic variants interact with nutritional factors to modify ADRD risk and 2) the mitochondrial peptide transcriptome is dramatically affected by high-fat diet in aged mice. We will center our focus by conducting cutting-edge genetic analyses to first see if certain mtSNPs attenuate or synergize with diet to affect ADRD risk. This project will engage a team of scientists to address variation in mitochondrial DNA (Aim 1) and mitochondrial RNA (Aim 2) with diet using the Health and Retirement Study cohort and RNASeq data on brain samples from mice fed a high-fat diet, respectively. Altogether, this supplement will identify mtDNA variants and potential peptides that interact with diet to modify ADRD risk byusing two independent approaches (well characterized large human databases and state-of-the- art novel transcriptomic approaches). If successful, findings from this project could lead to new target identification of peptides with therapeutic potential.